Boosting effect of high-dose influenza vaccination on innate immunity among elderly

Abstract

BACKGROUNDThe high-dose quadrivalent influenza vaccine (QIV-HD) showed superior efficacy against laboratory-confirmed illness compared with the standard-dose quadrivalent influenza vaccine (QIV-SD) in randomized controlled trials with the elderly. However, specific underlying mechanism remains unclear.METHODSThis phase IV randomized controlled trial compared early innate responses induced by QIV-HD and QIV-SD in 59 individuals aged > 65 years. Systemic innate cells and gene signatures at day 0 (D0) and D1 as well as hemagglutinin inhibition antibody (HIA) titers at D0 and D21 after vaccination were assessed.RESULTSQIV-HD elicited robust humoral response with significantly higher antibody titers and seroconversion rates than QIV-SD. At D1 after vaccination, QIV-HD recipients showed significant reduction in innate cells, including conventional DCs and NK cells, compared with QIV-SD, correlating with significantly increased HIA titers at D21. Blood transcriptomic analysis revealed greater amplitude of gene expression in the QIV-HD arm, encompassing genes related to innate immune response, IFNs, and antigen processing and presentation, and correlated with humoral responses. Interestingly, comparative analysis with a literature dataset from young adults vaccinated with influenza standard-dose vaccine highlighted strong similarities in gene expression patterns and biological pathways with the elderly vaccinated with QIV-HD.CONCLUSIONQIV-HD induces higher HIA titers than QIV-SD, a youthful boost of the innate gene expression significantly associated with high HIA titers.TRIAL REGISTRATIONEudraCT no. 2021-004573-32.

Keywords: Immunology; Influenza; Innate immunity; Vaccines.

Figure 1. Study protocol and immunogenicity of QIV-HD compared with QIV-SD in adults ≥ 65 years old.

(A) Schematic overview of experiment with kinetics of biological sample collection including serum collection and whole blood cells for transcriptomic and innate cell phenotype analyses. (B) Dot plots present geometric mean titers (GMT) with 95% CI of anti–A H1N1 hemagglutination inhibition antibody (HIA) titers, anti–A H3N2 HIA titers, anti–B Victoria HIA titers, and anti–B Yamagata HIA titers at D0, D21, D90, and D210 after QIV-SD (green) or QIV-HD (violet) vaccination. The horizontal line represents the seroprotection cut off with HIA titers at 40. Wilcoxon tests and Fisher’s exact tests were performed to compare the characteristics of the 2 groups (**P <0.01).

Figure 2. Differential expression genes by transcriptomic analysis at D1 after vaccination in QIV-HD and QIV-SD arms.

(A) Left panel: Volcano plot of differentially expressed genes (DEG) (D1/D0) of arm QIV-SD. Significance is defined by a FDR adjusted P < 0.05 and a FC > 1.41 (red; upregulated) or < –1.41 (blue; downregulated). Right panel: Volcano plot of DEG: (D1/D0) of arm QIV-HD. The vertical black lines delimit the 1.41-FC effects. (B) Differential FC (D1/D0) comparison of gene expression between QIV-SD and QIV-HD groups. The black square delimits the 1.41-FC effects. Significance is defined by an FDR adjusted P <0.05 and a FC > 1.41 (red; upregulated) or < –1.41 (blue; downregulated). (C) Biological processes enriched by the DEG in the QIV-HD. DEG were ranked by their value of FC (P <0.05), and the GSEA algorithm using Gene Ontology – Biological Processes was performed to extract functional information from gene expression data.

Figure 3. Early innate gene signature following QIV-HD vaccination in elderly.

(A) Heatmap of the log₂ FC values of the 213 genes correlated with the log₂ FC of anti–A H3N2 HIA titers. The graph displays the values for each patient, ordered by increasing values in the log₂ of antibody titers, which are displayed in the bar plot (Spearman correlation, P < 0.05 and r > 0.4). (B) Correlation plots of the log₂ FC of H3N2 HIA titers (y axis) and the log₂ FC of some of the top genes correlated with them (x axis). (C) STRING pathway analysis of genes correlated with the log₂ FC of anti–A H3N2 HIA titers, without disconnected nodes in the network. Edges represent protein-protein associations with minimum required interaction score of 0.4 edge confidence (line thickness indicates the strength of data support).

Figure 4. Innate cellular responses following QIV-HD and QIV-SD vaccination in elderly.

(A) Radar chart presents the 1 and –1 FC (log₂) for participant groups. The medians of each blood cell population (T and B cells; neutrophils; basophils; CD56^bright early and CD56^dim NK cells; classical, intermediate, and nonclassical monocytes; pDCs, cDC1, cDC2, and moDCs) are presented for QIV-SD (green) and QIV-HD (violet) groups. (B) Violin plot of log₂ FC (D1/D0) are represented for CD56^bright early NK cells and cDC2 for 2 vaccinated groups. (C) Principal component analysis (PCA) representation of QIV-SD (green) and QIV-HD (violet) groups based on induced blood cell populations at D1/D0. Arrows indicate the prominent parameter distinguishing features (neutrophils, CD56^bright early NK cells and cDC2). (D) Venn diagram of common DEG correlated with anti–A H3N2 (green) and –A H1N1 (yellow) HIA titers, CD56^bright early NK cell FC (blue), and cDC2 FC (red). (E) STRING pathway analysis (with 3 K-means clustering represented by 3 node colors) of 46 genes are in common between all these 4 parameters. Statistical analyses were performed with 1-way ANOVA and Bonferroni’s multiple-comparison tests, and indicated as adjusted P values. **P < 0.01, ****P < 0.0001.

Figure 5. Significant up- and downregulated gene associated with early innate response modules in elderly individuals vaccinated with QIV-SD and QIV-HD compared with young adults vaccinated with TIV-SD.

(A) The tmod enrichment analysis for QIV-SD, QIV-HD and young adults vaccinated with TIV-SD. From the list of genes provided by the limma R package, genes were ranked by their P values, and the enriched blood transcription modules were obtained by the CERNO test. The effect size is proportional to the size of the pie, while the adjusted P value is proportional to color intensity. Within each pie, the proportion of significantly upregulated and downregulated genes is shown in red and blue, respectively. The gray portion of the pie represents genes that are not significantly differentially regulated. (B) Heatmap of the log₂ FC of the biomarkers of immune response that belong to the IFN signaling pathway, comparing the QIV-SD and QIV-HD arms with the meta-analysis of studies with young adults vaccinated with TIV-SD. The combining approach from the MetavolcanoR package was used to summarize the FC of the studies based on the mean of the values, while the P value was combined using the Fisher method.